Hi
As promised you will find the instructions at: http://private.vlsi.informatik.tu-darmstadt.de/st/instructions.html
If you have any suggestions/corrections whatever please let me know.
I haven't had the time testing flashing the flash-roms, but i can switch between the two chips which have currently differen original bios versions on them.
Since i am going to start tinkering with linux bios real soon now, i would like to know if the instructions compiled by Ward are still the way to go?
ST
Hi ST,
On Wed, Apr 18, 2007 at 10:26:55PM +0200, ST wrote:
As promised you will find the instructions at: http://private.vlsi.informatik.tu-darmstadt.de/st/instructions.html
If you have any suggestions/corrections whatever please let me know.
Very nice!
I haven't had the time testing flashing the flash-roms, but i can switch between the two chips which have currently differen original bios versions on them.
Since i am going to start tinkering with linux bios real soon now, i would like to know if the instructions compiled by Ward are still the way to go?
They still work, but I've just verified that the mainline v2 tree now also works. So it's easier now :) I'm planning to update the tutorial soon; give me a couple of days.
Thanks, Ward.
Hi Ward
Very nice!
:-)
They still work, but I've just verified that the mainline v2 tree now also works. So it's easier now :) I'm planning to update the tutorial soon; give me a couple of days.
Ok. Current kernels only need "acpi_use_timer_override" as cmdline option for successfull booting.
ST
Hi ST,
On Wed, Apr 18, 2007 at 11:32:14PM +0200, ST wrote:
They still work, but I've just verified that the mainline v2 tree now also works. So it's easier now :) I'm planning to update the tutorial soon; give me a couple of days.
Ok. Current kernels only need "acpi_use_timer_override" as cmdline option for successfull booting.
Indeed! I've just verified that. I'll update the tutorial.
Thanks, Ward.
On Wed, Apr 18, 2007 at 11:32:14PM +0200, ST wrote:
Current kernels only need "acpi_use_timer_override" as cmdline option for successfull booting.
Do we know why?
//Peter
On Wed, Apr 18, 2007 at 10:26:55PM +0200, ST wrote:
http://private.vlsi.informatik.tu-darmstadt.de/st/instructions.html
Shows up blank in my mozilla, but read it in lynx.
If you have any suggestions/corrections whatever please let me know.
What size is the SMD resistor? 0805, 0603 or something else? And what does it do? Where is it connected?
I haven't had the time testing flashing the flash-roms, but i can switch between the two chips which have currently differen original bios versions on them.
Please confirm once you've tried flashrom. Booting only requires reads to work, but IIRC all relevant signals were connected across.
//Peter
Hi Peter
Am Donnerstag, 19. April 2007 05:44 schrieb Peter Stuge:
On Wed, Apr 18, 2007 at 10:26:55PM +0200, ST wrote:
http://private.vlsi.informatik.tu-darmstadt.de/st/instructions.html
Shows up blank in my mozilla, but read it in lynx.
Huh?! Thats really plain HTML and i could read it with Konqui and Iceweasel (aka Firefox). But to be sure i just validated the page: http://validator.w3.org/check?uri=http%3A%2F%2Fprivate.vlsi.informatik.tu-da...
What size is the SMD resistor? 0805, 0603 or something else? And what does it do? Where is it connected?
The resistor is connected to INIT# of PLCC32 socket. It's an pullup which has been missing for the open pads. This pullup will eventually be pulled down by the switch connected to pad T12.
Please confirm once you've tried flashrom. Booting only requires reads to work, but IIRC all relevant signals were connected across.
Yes but first i have to get some rom image. Or is the original firmware from Gigabyte in a format that can be used by flashbios?
Current kernels only need "acpi_use_timer_override" as cmdline option for successfull booting.
Do we know why?
Well actually i filed a bugreport: http://bugzilla.kernel.org/show_bug.cgi?id=8219
ST
ST wrote:
Am Donnerstag, 19. April 2007 05:44 schrieb Peter Stuge:
--snip--
Please confirm once you've tried flashrom. Booting only requires reads to work, but IIRC all relevant signals were connected across.
Yes but first i have to get some rom image. Or is the original firmware from Gigabyte in a format that can be used by flashbios?
Yes, flashrom can use any (uncompressed, unencrypted) bios image, including the factory one.
-Corey
Hi,
On Thu, Apr 19, 2007 at 10:33:13AM +0200, ST wrote:
Shows up blank in my mozilla, but read it in lynx.
Huh?! Thats really plain HTML and i could read it with Konqui and Iceweasel (aka Firefox). But to be sure i just validated the page:
Now it displays properly. Probably mozilla gone crazy.
What size is the SMD resistor? 0805, 0603 or something else?
This is important. It seems this would be a 0603 resistor. Please add that to the page.
And what does it do? Where is it connected?
The resistor is connected to INIT# of PLCC32 socket. It's an pullup which has been missing for the open pads.
Ok! Perhaps add that under a "Details of operation" section on the page too.
This pullup will eventually be pulled down by the switch connected to pad T12.
Yep, nice work.
Please confirm once you've tried flashrom. Booting only requires reads to work, but IIRC all relevant signals were connected across.
Yes but first i have to get some rom image. Or is the original firmware from Gigabyte in a format that can be used by flashbios?
I don't think flashrom cares at all about the contents.
Current kernels only need "acpi_use_timer_override" as cmdline option for successfull booting.
Do we know why?
Well actually i filed a bugreport: http://bugzilla.kernel.org/show_bug.cgi?id=8219
Nice! Still status NEEDINFO though. :\
//Peter
Hi
What size is the SMD resistor? 0805, 0603 or something else?
This is important. It seems this would be a 0603 resistor. Please add that to the page.
Ok, added.
And what does it do? Where is it connected?
The resistor is connected to INIT# of PLCC32 socket. It's an pullup which has been missing for the open pads.
Ok! Perhaps add that under a "Details of operation" section on the page too.
Added.
Yes but first i have to get some rom image. Or is the original firmware from Gigabyte in a format that can be used by flashbios?
I don't think flashrom cares at all about the contents.
Don't know about flashbios. In the good old EEPROM times there was s.t. like intel hex format or so and some writers only took this format. I'll try writing later on.
Current kernels only need "acpi_use_timer_override" as cmdline option for successfull booting.
Do we know why?
Well actually i filed a bugreport: http://bugzilla.kernel.org/show_bug.cgi?id=8219
Nice! Still status NEEDINFO though. :\
Well, i provided the info, but it somehow got lost on the radar? I didn't dare to change the Status.
Regards ST
Hi
Please confirm once you've tried flashrom. Booting only requires reads to work, but IIRC all relevant signals were connected across.
Yesterday evening i tested writing to flash with flashrom. Albeit still only with the original firmware blobs in different versions to verify that s.t. had changed.
I tested writing to both chips and switching chips while running. Thus i verified that: * only one chip is written at a time, * both chips are writeable, * it is possible to switch the flash-rom's while running.
So i think everything except the question about the direct ground pulldown of the INIT# pins (with or without resistor) is clear that it works as expected.
Concerning the pulldown: I currently have only a jumper connected to the chips, this makes switching somehow delicate since it takes some time pull the jumper of and put it on the other location. In this time both chips are active, which is not to nice, but i can report that it works nevertheless. I had some suggestions to put some resistors in between the switch and the T12 and T22 pins.
I also tried to build linuxbios with a linux payload, but somehow the kernel bzImage was way to big to fit into the flash. Is it possible to fit a 2.6.2x kernel into 4MBit flash-rom?
Thanks ST
On Fri, Apr 20, 2007 at 10:06:57AM +0200, ST wrote:
I tested writing to both chips and switching chips while running. Thus i verified that:
- only one chip is written at a time,
- both chips are writeable,
- it is possible to switch the flash-rom's while running.
Nice!
So i think everything except the question about the direct ground pulldown of the INIT# pins (with or without resistor) is clear that it works as expected.
Concerning the pulldown: I currently have only a jumper connected to the chips
Ideally you would have both INIT# pins pulled down instead of up, and then connect either of the two pins to the chipset INIT# output.
I also tried to build linuxbios with a linux payload, but somehow the kernel bzImage was way to big to fit into the flash. Is it possible to fit a 2.6.2x kernel into 4MBit flash-rom?
That will be very tight, probably not possible without a lot of work. (Maybe try out linux-tiny.)
But since you have a socket and are using LPC flash, you could try a 8Mbit (or even 16Mbit) flash chip and see if your chipset decodes that large a window to the flash.
//Peter
Hi ST, nice work, on this days I will try it an take some pictures too, I want put the switch externally of the case for easing switching between the bios.
Bye.
On 4/18/07, ST st@iss.tu-darmstadt.de wrote:
Hi
As promised you will find the instructions at: http://private.vlsi.informatik.tu-darmstadt.de/st/instructions.html
If you have any suggestions/corrections whatever please let me know.
I haven't had the time testing flashing the flash-roms, but i can switch between the two chips which have currently differen original bios versions on them.
Since i am going to start tinkering with linux bios real soon now, i would like to know if the instructions compiled by Ward are still the way to go?
ST
-- linuxbios mailing list linuxbios@linuxbios.org http://www.linuxbios.org/mailman/listinfo/linuxbios